enable archive location in crossref metadata

front-matter · Nov 4, 2023 · 0dd5f8e · 0dd5f8e
1 parent f64fa7e
commit 0dd5f8e
Show file tree

Hide file tree

Showing 11 changed files with 307 additions and 34 deletions.
diff --git a/Gemfile.lock b/Gemfile.lock
@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    commonmeta-ruby (3.6.1)
+    commonmeta-ruby (3.7.0)
       activesupport (>= 4.2.5, < 8.0)
       addressable (~> 2.8.1, < 2.8.2)
       base32-url (>= 0.7.0, < 1)
@@ -27,7 +27,7 @@ PATH
 GEM
   remote: https://rubygems.org/
   specs:
-    activesupport (7.1.0)
+    activesupport (7.1.1)
       base64
       bigdecimal
       concurrent-ruby (~> 1.0, >= 1.0.2)
@@ -128,7 +128,7 @@ GEM
       ffi-compiler (~> 1.0)
       rake (~> 13.0)
     logutils (0.6.1)
-    loofah (2.21.3)
+    loofah (2.21.4)
       crass (~> 1.0.2)
       nokogiri (>= 1.12.0)
     matrix (0.4.2)
@@ -151,12 +151,12 @@ GEM
     props (1.2.0)
       iniparser (>= 0.1.0)
     public_suffix (4.0.7)
-    racc (1.7.1)
+    racc (1.7.3)
     rack (3.0.8)
     rack-test (2.1.0)
       rack (>= 1.3)
     rainbow (3.1.1)
-    rake (13.0.6)
+    rake (13.1.0)
     rdf (3.3.1)
       bcp47_spec (~> 0.2)
       link_header (~> 0.0, >= 0.0.8)
@@ -171,7 +171,7 @@ GEM
     rdf-xsd (3.3.0)
       rdf (~> 3.3)
       rexml (~> 3.2)
-    regexp_parser (2.8.1)
+    regexp_parser (2.8.2)
     rexml (3.2.6)
     rspec (3.12.0)
       rspec-core (~> 3.12.0)
@@ -189,19 +189,18 @@ GEM
     rspec-xsd (0.1.0)
       nokogiri (~> 1.6)
       rspec (~> 3)
-    rubocop (1.56.4)
-      base64 (~> 0.1.1)
+    rubocop (1.57.2)
       json (~> 2.3)
       language_server-protocol (>= 3.17.0)
       parallel (~> 1.10)
-      parser (>= 3.2.2.3)
+      parser (>= 3.2.2.4)
       rainbow (>= 2.2.2, < 4.0)
       regexp_parser (>= 1.8, < 3.0)
       rexml (>= 3.2.5, < 4.0)
       rubocop-ast (>= 1.28.1, < 2.0)
       ruby-progressbar (~> 1.7)
       unicode-display_width (>= 2.4.0, < 3.0)
-    rubocop-ast (1.29.0)
+    rubocop-ast (1.30.0)
       parser (>= 3.2.1.0)
     rubocop-capybara (2.19.0)
       rubocop (~> 1.41)
@@ -212,8 +211,8 @@ GEM
       rubocop-ast (>= 0.4.0)
     rubocop-rake (0.6.0)
       rubocop (~> 1.0)
-    rubocop-rspec (2.24.1)
-      rubocop (~> 1.33)
+    rubocop-rspec (2.25.0)
+      rubocop (~> 1.40)
       rubocop-capybara (~> 2.17)
       rubocop-factory_bot (~> 2.22)
     ruby-progressbar (1.13.0)
@@ -236,7 +235,7 @@ GEM
       logutils (>= 0.6.1)
       props (>= 1.1.2)
       rubyzip (>= 1.0.0)
-    thor (1.2.2)
+    thor (1.3.0)
     tzinfo (2.0.6)
       concurrent-ruby (~> 1.0)
     unf (0.1.4)

diff --git a/lib/commonmeta/crossref_utils.rb b/lib/commonmeta/crossref_utils.rb
@@ -65,6 +65,7 @@ def insert_journal(xml)
           insert_funding_references(xml)
           insert_crossref_access_indicators(xml)
           insert_crossref_relations(xml)
+          insert_archive_locations(xml)
           insert_doi_data(xml)
           insert_citation_list(xml)
         end
@@ -85,6 +86,7 @@ def insert_posted_content(xml)
         insert_funding_references(xml)
         insert_crossref_access_indicators(xml)
         insert_crossref_relations(xml)
+        insert_archive_locations(xml)
         insert_doi_data(xml)
         insert_citation_list(xml)
       end
@@ -327,6 +329,16 @@ def insert_item_number(xml)
       end
     end
 
+    def insert_archive_locations(xml)
+      return xml if archive_locations.blank?
+
+      xml.archive_locations do
+        archive_locations.each do |archive_location|
+          xml.archive("name" => archive_location)
+        end
+      end
+    end
+
     def insert_doi_data(xml)
       return xml if doi_from_url(id).blank? || url.blank?
 

diff --git a/lib/commonmeta/metadata.rb b/lib/commonmeta/metadata.rb
@@ -10,7 +10,7 @@ class Metadata
                   :email, :registrant
     attr_reader :doc, :page_start, :page_end
     attr_writer :id, :provider_id, :client_id, :doi, :alternate_identifiers, :contributors,
-                :titles, :publisher, :license, :date, :volume, :url, :version, :subjects, :descriptions, :language, :sizes, :formats, :schema_version, :meta, :container, :provider, :format, :funding_references, :state, :geo_locations, :type, :additional_type, :files, :references, :related_identifiers, :related_items, :style, :locale
+                :titles, :publisher, :license, :date, :volume, :url, :version, :subjects, :descriptions, :language, :sizes, :formats, :schema_version, :meta, :container, :provider, :format, :funding_references, :state, :geo_locations, :type, :additional_type, :files, :references, :related_identifiers, :related_items, :style, :locale, :archive_locations
 
     def initialize(options = {})
       options.symbolize_keys!
@@ -253,5 +253,9 @@ def titles
     def contributors
       @contributors ||= meta.fetch('contributors', nil)
     end
+
+    def archive_locations
+      @archive_locations ||= meta.fetch('archive_locations', nil)
+    end
   end
 end
diff --git a/lib/commonmeta/readers/json_feed_reader.rb b/lib/commonmeta/readers/json_feed_reader.rb
@@ -69,6 +69,19 @@ def read_json_feed_item(string: nil, **options)
         funding_references = get_funding_references(meta)
         related_identifiers = get_related_identifiers(meta)
         alternate_identifiers = [{ "alternateIdentifier" => meta["id"], "alternateIdentifierType" => "UUID" }]
+
+        archive_url = meta.fetch("archive_url", nil)
+
+        if archive_url
+          host = URI.parse(archive_url).host
+          if ["web.archive.org", "wayback.archive-it.org"].include?(host)
+            archive_locations = ["Internet Archive"]
+          else
+            archive_locations = nil
+          end
+        else
+          archive_locations = nil
+        end
 
         { "id" => id,
           "type" => type,
@@ -86,6 +99,7 @@ def read_json_feed_item(string: nil, **options)
           "funding_references" => funding_references.presence,
           "related_identifiers" => related_identifiers.presence,
           "alternate_identifiers" => alternate_identifiers,
+          "archive_locations" => archive_locations,
           "state" => state }.compact.merge(read_options)
       end
 

diff --git a/lib/commonmeta/schema_utils.rb b/lib/commonmeta/schema_utils.rb
@@ -5,7 +5,7 @@
 
 module Commonmeta
   module SchemaUtils
-    COMMONMETA = File.read(File.expand_path("../../resources/commonmeta_v0.10.3.json",
+    COMMONMETA = File.read(File.expand_path("../../resources/commonmeta_v0.10.4.json",
                                             __dir__))
 
     def json_schema_errors

diff --git a/lib/commonmeta/version.rb b/lib/commonmeta/version.rb
@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Commonmeta
-  VERSION = '3.6.1'
+  VERSION = '3.7.0'
 end
diff --git a/resources/commonmeta_v0.10.3.json → resources/commonmeta_v0.10.4.json b/resources/commonmeta_v0.10.3.json → resources/commonmeta_v0.10.4.json
@@ -1,7 +1,7 @@
 {
   "$schema": "http://json-schema.org/draft-07/schema#",
-  "$id": "https://commonmeta.org/commonmeta_v0.10.3.json",
-  "title": "Commonmeta v0.10.3",
+  "$id": "https://commonmeta.org/commonmeta_v0.10.4.json",
+  "title": "Commonmeta v0.10.4",
   "description": "JSON representation of the Commonmeta schema.",
   "additionalProperties": false,
   "definitions": {
@@ -516,6 +516,15 @@
       "description": "The state of the resource.",
       "type": "string",
       "enum": ["findable", "not_found"]
+    },
+    "archive_locations": {
+      "description": "The location where content is archived.",
+      "type": "array",
+      "items": {
+        "type": "string",
+        "enum": ["CLOCKSS", "Internet Archive", "KB", "LOCKSS", "Portico", "Other"]
+      },
+      "uniqueItems": true
     }
   },
   "required": [

diff --git a/...vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/archived_wordpress_post.yml b/...vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/archived_wordpress_post.yml